As people in the US live longer, persons with comorbid Alzheimer disease or related dementias (hereafter dementia) and diabetes become increasingly common in clinical care. Diabetes is a largely ambulatory and self-managed condition that should not require inpatient care. Persons living with dementia and diabetes (PWD&D), however, have high risks of hospitalizations. These hospitalizations are considered preventable and potentially avoidable. There is a need to understand care considerations in PWD&D that predispose them to potentially preventable hospitalizations (PPH), and this has been articulated as an important national goal. In diabetes, glycemic control has been fundamental in clinical care as it reduces risks of diabetes complications. Older adults with diabetes, in particular those with cognitive deficits, however, may have difficulty managing glucose at home and therefore have higher rates of diabetes complications. PWD&D are also susceptible to overtreatment and have increased health care utilization because of hypoglycemia. Therefore, both the risks of diabetes complications and the risks of therapy (e.g., hypoglycemia) should be considered when setting therapeutic goals. However, the most appropriate glycemic targets for these patients have not been empirically determined. The AGS and ADA have promulgated recommendations on optimal glucose targets in persons with cognitive impairment, but these are consensus statements and should be strengthened through higher levels of evidence. To our knowledge, no prior prospective study with research standards for dementia diagnosis has evaluated the implications of incident dementia for glucose control among people with diabetes and, particularly in terms of rates of hospitalizations, PPH, and diabetes-related PPH. We propose to examine the relationship between dementia onset, glucose levels, and rates of different types of hospitalizations in people with diabetes from the NIA-funded Adult Changes in Thought (ACT) study. ACT resources include extensive clinical laboratory data, incident research-quality dementia diagnoses and complete capture of hospitalization and medication data. Our first aim will be to determine the extent to which dementia onset is associated with higher rates of overall hospitalizations, rates of PPH and rates of diabetes-related PPH among persons with treated diabetes. Our second aim will be to evaluate the association of average time-varying individual-level glucose levels with the rates of overall hospitalization, PPH, and diabetes-related PPH while accounting for incident dementia and other time-varying confounders. By leveraging ACT, the proposed study provides an extraordinary opportunity for new observational insights into the associations among dementia, glucose control and hospitalizations in older persons with diabetes. Findings from this study address national priority and will advance our understanding of the role of glycemic control on admission risks in an understudied aged population and will provide preliminary data for developing strategies that mitigate the adverse effects of over- or under- treatment in diabetes management.